We Won't Win Because We're Smart. We'll Win Because We're Crazy.

Palantir's CTO on heretics, the industrial base, and why every real breakthrough happens despite the system, never because of it.

Shyam Sankar says something at the end that reframes the whole conversation. Asked for the case for American optimism, he doesn't cite GDP or chips or the military budget. He says: the reason we'll win in any domain is that we're crazy. We don't even know what we'll do when we get punched in the face. That unpredictability — that plasticity — is the actual advantage. It's a strange thing for the CTO of a defense company to lead with, and that's exactly why it's worth sitting with.

The spine of his argument is a single, uncomfortable pattern: almost no defense innovation in history happened because of the system. All of it happened despite the system. He calls the people who pulled it off heretics. Churchill built the tank as a rogue project when the British Army said it had horses and didn't need "land ships." Andrew Higgins, a foul-mouthed boat builder in Louisiana, got shut out by the Navy — then built the landing craft that 92% of WWII boats were based on. No Higgins boats, no Normandy, no Iwo Jima. Admiral Rickover built the nuclear navy after Oppenheimer himself told him it wouldn't work. The through-line isn't genius. It's someone refusing to accept a "no" that the entire institution had already agreed on.

Here's the part I keep coming back to, because it's a management truth disguised as a war story: heretics need two things, and only two. Space for the heresy to exist, and a modicum of protection from the top. That's it. Sankar's point is that we're not short on heretics — we have them, we always have them. The organism just rejects them by default, the way a body rejects a transplant. Every company, every institution, quietly expels its most talented people because they're inconvenient. The job of leadership isn't to generate genius. It's to shield it long enough to survive contact with the bureaucracy. Palantir itself was a pack of heretics who got beaten by the mainstream for years — and the chip on the shoulder, the need to be ten times better than the alternative, turned out to be the fuel.

The strategic insight underneath the war stories is about deterrence, and it's a correction most people get backwards. We think the stockpile is the deterrent. It isn't. The ability to generate the stockpile is. Ukraine burned through ten years of production in ten weeks of fighting — and the lesson isn't "build a bigger pile," it's that we optimized ourselves into fragility. We chased efficiency over effectiveness until every weapon got more exquisite, more expensive, and therefore rarer. In WWII the Germans were the superior engineers building small, perfect quantities. We just produced more, and it won. Today we've become the Germans, and China has become who we used to be. That inversion should terrify anyone paying attention, and fixing it is less about a breakthrough than about a mindset — get to effectiveness first, then worry about efficiency, never the other way around.

And on the question everyone actually wants answered — AI weapons, killer robots, humans out of the loop — Sankar's move is to deflate the drama. We've had autonomous weapons since the '70s. The Aegis system defending a carrier fires on incoming missiles faster than any human could approve, because the ship would already be hit. A missile fired from a jet closes its own kill chain. So this is a difference of degree, not of kind — and because we keep miscategorizing it as a total difference of kind, we end up having deranged conversations about it. His warning cuts the other way too, though, through the story of Theodore Hall: the eighteen-year-old Manhattan Project genius who handed the Soviets the bomb, not out of malice but because he assumed being brilliant at physics made him brilliant at geopolitics. Tyranny by tech bro. The lesson isn't "trust the smart people." It's that the smartest people most need epistemic humility — and that in a democracy, the ones making these calls should be accountable to the public, not to their own certainty.

The line I'll keep: America's greatest risk isn't homicide. It's suicide. Nobody beats us. We only beat ourselves — by forgetting what made us worth betting on, by rejecting our own heretics, by optimizing the craziness out of the system that was the whole point.